HEALTH SAFETY, AND ENVIRONMENT

advertisement
HEALTH a nd INFRASTRUCTURE,
SAFETY, AND ENVIRONMENT
CHILDREN AND FAMILIES
EDUCATION AND THE ARTS
The RAND Corporation is a nonprofit institution that helps improve policy and
decisionmaking through research and analysis.
ENERGY AND ENVIRONMENT
HEALTH AND HEALTH CARE
INFRASTRUCTURE AND
TRANSPORTATION
This electronic document was made available from www.rand.org as a public service
of the RAND Corporation.
INTERNATIONAL AFFAIRS
LAW AND BUSINESS
Skip all front matter: Jump to Page 16
NATIONAL SECURITY
POPULATION AND AGING
PUBLIC SAFETY
SCIENCE AND TECHNOLOGY
TERRORISM AND
HOMELAND SECURITY
Support RAND
Browse Reports & Bookstore
Make a charitable contribution
For More Information
Visit RAND at www.rand.org
ExploreRAND Health
RAND Infrastructure, Safety, and Environment
View document details
Limited Electronic Distribution Rights
This document and trademark(s) contained herein are protected by law as indicated in a notice appearing
later in this work. This electronic representation of RAND intellectual property is provided for noncommercial use only. Unauthorized posting of RAND electronic documents to a non-RAND website is
prohibited. RAND electronic documents are protected under copyright law. Permission is required from
RAND to reproduce, or reuse in another form, any of our research documents for commercial use. For
information on reprint and linking permissions, please see RAND Permissions.
This product is part of the RAND Corporation technical report series. Reports may
include research findings on a specific topic that is limited in scope; present discussions
of the methodology employed in research; provide literature reviews, survey instruments, modeling exercises, guidelines for practitioners and research professionals, and
supporting documentation; or deliver preliminary findings. All RAND reports undergo rigorous peer review to ensure that they meet high standards for research quality
and objectivity.
TECHNIC A L REP O RT
National Evaluation of Safe Start
Promising Approaches
Results Appendix I: Multnomah County,
Oregon
In Jaycox, L. H., L. J. Hickman, D. Schultz, D. Barnes-Proby, C. M. Setodji, A. Kofner, R. Harris, J. D.
Acosta, and T. Francois, National Evaluation of Safe Start Promising Approaches: Assessing Program
Outcomes, Santa Monica, Calif.: RAND Corporation, TR-991-1-DOJ, 2011
Sponsored by the U.S. Department of Justice’s Office of Juvenile Justice and Delinquency Prevention
HEA LTH and
INFRASTRUCTURE, SAFETY, A N D EN VI R ON MEN T
This research was sponsored by the U.S. Department of Justice’s Office of Juvenile Justice
and Delinquency Prevention and was conducted under the auspices of the Safety and
Justice Program within RAND Infrastructure, Safety, and Environment and under RAND
Health’s Health Promotion and Disease Prevention Program.
Library of Congress Control Number: 2011935596
ISBN: 978-0-8330-5822-5
The R AND Corporation is a nonprofit institution that helps improve policy and
decisionmaking through research and analysis. RAND’s publications do not necessarily
reflect the opinions of its research clients and sponsors.
R® is a registered trademark.
© Copyright 2011 RAND Corporation
Permission is given to duplicate this document for personal use only, as long as it
is unaltered and complete. Copies may not be duplicated for commercial purposes.
Unauthorized posting of RAND documents to a non-RAND website is prohibited. RAND
documents are protected under copyright law. For information on reprint and linking
permissions, please visit the RAND permissions page (http://www.rand.org/publications/
permissions.html).
Published 2011 by the RAND Corporation
1776 Main Street, P.O. Box 2138, Santa Monica, CA 90407-2138
1200 South Hayes Street, Arlington, VA 22202-5050
4570 Fifth Avenue, Suite 600, Pittsburgh, PA 15213-2665
RAND URL: http://www.rand.org
To order RAND documents or to obtain additional information, contact
Distribution Services: Telephone: (310) 451-7002;
Fax: (310) 451-6915; Email: order@rand.org
MULTNOMAH COUNTY, OREGON, SAFE START OUTCOMES REPORT
ABSTRACT
The Multnomah County Safe Start program developed an intervention for
young children (birth to 6 years of age) exposed to domestic violence and
involved with child welfare services. The program is fully described in National
Evaluation of Safe Start Promising Approaches: Assessing Program Implementation
(Schultz et al., 2010). The program planned a quasi-experimental trial comparing
similar children in the different child welfare offices within Multnomah County.
The Safe Start program staff enrolled 43 families in the study, 32 in the
intervention group and 11 in the comparison. By the time of the first follow-up
assessment, only 11 families remained in the Multnomah County intervention
group and five in the comparison group. Because of low overall study
participation, enrollment ended about two months early, and follow-up of
enrolled families ended about eight months before other sites.
At baseline, caregivers reported that children had been exposed to an
average of 2.9 types of violence in their lives. Twenty-three percent of caregivers
in the overall sample reported child posttraumatic stress disorder (PTSD)
symptoms in the “significant” range, and 33 percent reported total parenting
stress levels in the “clinical” range. Overall, 63 percent of the intervention group
families received at least one session with a Multnomah County advocate and
received at least one case coordination meeting. Child-Parent Psychotherapy
(CPP) was provided to 28 percent of the intervention group families, with an
average number of 16 sessions per family. Of the 11 intervention group families
in the retained sample at six months, eight (73 percent) received at least one
session with a Safe Start advocate, with an average of 23 sessions. Eight families
who received case coordination services and three families in the retained
sample participated in at least one CPP therapy service.
Evaluation of the impact of the intervention was not possible because of a
lack of a comparison group to examine difference in outcomes over time. Limited
outcome analyses were conducted to examine changes between baseline and the
six-month assessment within the intervention group only. In these analyses, no
significant changes in outcomes were observed, with the exception of caregiver
1
experience of domestic violence. On this measure, however, a decrease would be
expected because the baseline observation period is considerably longer than the
six-month window covered by the first follow-up research assessment.
Overall, the sample size limitations mean that no conclusions can be
drawn about the impact of the Multnomah County Safe Start intervention as
implemented on child- and family-level outcomes. However, our process
evaluation found that the program model did appear to have important impacts
in terms of increasing integration and coordination of services between child
welfare and domestic violence advocacy service providers (Schultz et al., 2010).
This is a positive finding, as system-level change was a key goal of the program.
The intervention model’s impact on individual outcomes, however, remains
unknown, and conclusions await testing with an adequate sample size.
INTRODUCTION
The Multnomah County Safe Start program is located in Multnomah
County, Oregon, its most populous county. Safe Start intervention services were
provided in Gresham, a city adjacent to Portland. Compared to the rest of the
state, Multnomah County reported a higher risk of co-occurring domestic
violence exposure among its child abuse cases managed by child welfare services
(37 percent in Multnomah County cases relative to 28 percent statewide)
(Multnomah County, 2004). The Safe Start program was designed to address a
deficiency in coordinated services for the issues of child abuse and domestic
violence, as well as poor communication between the two systems that respond
these problems (Multnomah County, 2004). Thus, the primary focus of the
program was to develop an intervention for young children exposed to domestic
violence and involved with child welfare services.
According to the Safe Start program’s original Safe Start proposal to the
U.S. Department of Justice’s Office of Juvenile Justice and Delinquency
Prevention, the two targeted service systems operated independently to meet
their respective objectives, with child needs being the primary focus of child
welfare services and adult victim needs being the main thrust of domestic
violence services (Multnomah County, 2004). Consequently, the relationship
between child welfare agencies and domestic violence programs has been
strained at times, and even sometimes described as incompatible (Aron and
Olson, 1997). The Multnomah County program is part of an innovation in
2
thinking about how services should be delivered to children involved in child
welfare services who have been exposed to domestic violence. This trend focuses
on seeking to integrate the missions and practices of child welfare and domestic
services providers such that they would be working together toward joint goals
rather than at cross-purposes. An example of prior efforts that influenced the
development of the Multnomah County program was developed by the
Massachusetts Department of Social Services (Schechter, 1994). This model colocated domestic violence and child welfare services under one roof and
prioritized collaborative service provision. Similarly, in the Multnomah County
Safe Start intervention, direct services were provided through the coordinated
efforts of both the child welfare and domestic violence service systems, with the
explicit intention of ameliorating the negative effects of violence exposure on
child emotional well-being and the caregiver-child relationship. Thus, the
Multnomah County Safe Start project represents a relatively new and creative
approach to delivering services to meet the needs of both child welfare–involved
children and adult victims of domestic violence.
Because of low enrollment and retention in the Multnomah County Safe
Start program, the outcomes evaluation detailed here presents descriptive
information about baseline observations only.
•
•
•
•
•
•
MULTNOMAH COUNTY SAFE START
Intervention type: Domestic violence advocacy, CPP, and case
coordination and consultation
Intervention length: For all components, flexible depending on meeting
goals of case plan
Intervention setting: In-home and office-based
Target population: Children exposed to domestic violence within a county
child welfare population
Age range: 0–6
Primary referral source: Gresham Branch Office of County Child Welfare
Services
3
INTERVENTION
The Multnomah County Safe Start program involved three main
components: domestic violence advocacy, a modified version of CPP, and case
coordination and consultation. For all components, the intervention length
varied depending on individual need and case plan. The services were voluntary
and offered in addition to usual services and procedures of the Gresham Child
Welfare Office. Services were provided at a location most convenient for the
client, including in the home. Domestic violence (DV) advocacy was intended to
be the primary service. Case coordination and consultation among the DV
advocates, child welfare workers, and (when applicable) the CPP provider was
also expected for all clients. A modified version of CPP would be offered as an
additional service for mothers who expressed an interest in improving their
parenting or parent-child relationship. These components are described briefly in
the following paragraphs. For a full description of the Multnomah County
intervention as it was delivered, see Schultz et al. (2010).
The domestic violence advocates were physically located in the Gresham
Child Welfare Office to facilitate the identification of cases and to facilitate direct
service coordination with child welfare workers. The initial advocacy services
involved conducting a domestic violence–focused safety assessment. The
advocates also assessed whether the family’s basic needs were being met for such
things as food, clothing, housing, and utilities. The advocates then worked with
the mother to develop a safety plan and assisted her as needed to meet the basic
needs of herself and her children, in collaboration with child welfare. The
advocates also offered adult domestic violence victim support groups and
provided individual social support, such as accompanying mothers to court
hearings. Ultimately, the model expected that domestic violence–specific
advocacy services would improve the mother’s life circumstances and
functioning, which would in turn improve those of the children.
The therapy component involved a modification of the Lieberman and
Van Horn model for CPP (Lieberman, Van Horn, and Ghosh Ippen, 2005). In
Multnomah County, the CPP was delivered in clients’ homes, where sessions
were expected to last 60 to 90 minutes. Using “Don’t Hit My Mommy” (Lieberman
and Van Horn, 2005) as a procedural guide, the therapist focused on helping
mothers recognize, understand, and respond to the impact of domestic violence
on their children and on their own parenting. The intake into the counseling
4
component involved taking information about the family’s domestic violence
history; reviewing the mother’s concerns about her parenting, her relationship
with her children, and their behaviors; and developing a case plan with the
mother setting out goals (tied to the mother’s specific concerns) to work toward
over the course of the sessions. The sessions ended when the goals of case plan
were achieved.
In addition to the direct services provided to families, the Multnomah
County Safe Start program provided case coordination services, which included
discussions and joint case planning between child welfare workers and domestic
violence advocates coordinate services to families they are jointly serving. Case
coordination also involved both formal case review meetings (which could
include multiple other service providers, such as the parent-child specialist) and
informal conversations as the child welfare workers and advocates interacted
with one another around the office or stopped by each other’s work spaces to
“touch base” about a particular family. Consultation activities, such as crosstraining, technical assistance, and monthly partner meetings, allowed both
advocates and child welfare workers to gain a better understanding of each
other’s approach to working with individual families, including best practices
and legal and agency policy requirements.
Efforts to ensure and monitor the quality of the program included the
provision of specific training on the intervention, the child welfare system, and
implementation plans early in the project. In addition, the domestic violence
advocates’ services were monitored through monthly meetings and as-needed
consultation with their respective agency supervisors, and the advocates
participated in the monthly partner meetings to discuss their service provision
overall. The parent-child specialist received clinical supervision once per week
through her agency, augmented by once-weekly meetings with her direct
supervisor to discuss her services and individual cases. Also, the domestic
violence advocates, parent-child specialist, and child welfare workers had prior
experience working with victims of domestic violence.
5
METHOD
Design Overview
The design of this study was a quasi-experimental effectiveness trial
comparing children in the Gresham Child Welfare Office with similar children
recruited from three other child welfare branch offices within Multnomah
County. The treatment group from the Gresham office received domestic
violence advocacy, case coordination, and a modified version of CPP. Study
enrollees from the comparison branches of county child welfare offices received
child welfare case management services and referrals as usual. The planned data
collection was to assess child outcomes and contextual information at baseline
and at six, 12, 18, and 24 months. As discussed below, however, no assessments
could be completed after the six-month mark. Study enrollment took place
between January 2007 and March 2009 (over 27 months).
Evaluation Eligibility Criteria
Eligible domestic violence victims with concurrent child abuse and/or
neglect issues in their families were invited to participate in the Safe Start
program. Inclusion criteria for Safe Start services were as follows: (1) Current or
recent domestic violence issues were identified as a reason for child welfare
referral or as a family issue during the subsequent child welfare investigation; (2)
the mother (or primary female caregiver) of the children in the case had
experienced domestic violence; (3) at least one child in the family was 6 years old
or younger; and (4) the family had proficiency in English or Spanish.
When more than one child in the eligible age range was eligible for the
program by virtue of exposure to violence, the child with the closest birth date to
the date of enrollment was selected as the target child for the evaluation.
Recruitment of the Treatment and Comparison Groups
Because of the child welfare context of the research, recruitment into the
study for both the treatment and comparison groups had to proceed in a twostage process. In Oregon, researchers are prohibited from directly contacting
families identified in child welfare records. Only authorized personnel within the
Oregon Department of Human Services (child welfare) can contact families to
ask if they are interested in hearing about research participation. Once this
6
“consent to contact” is obtained, researchers can proceed to approach families
about study participation. For treatment group recruitment, the Safe Start
Domestic Violence Advocate stationed in the Gresham office was authorized to
seek consent to contact from eligible families. In the comparison group offices,
however, Safe Start program staff were not available, so designated child welfare
personnel filled this role.
To identify eligible children for the treatment group, Gresham Protective
Services supervisors looked for cases that met study eligibility criteria as part of
their standard procedures for reviewing new referrals to the Child Welfare Office.
The supervisors would then provide the case information to the Safe Start
domestic violence advocate. The advocate would then seek to contact the
primary caregiver, explain the available Safe Start services, confirm study
eligibility, and introduce the study. For caregivers willing to hear more about
potential study participation, contact information was provided to the research
team. The research team would then seek to set up a meeting with the caregiver
to fully explain the study, obtain consent for participation, and complete the
process of enrolling the family in the study. Caregivers could refuse study
participation and still receive all Safe Start services. In fact, the program sought
to keep confidential from the service providers the identities of the families that
did choose to enroll in the study.
For the comparison group, an equivalent procedure for identifying
potentially eligible cases was not feasible in the comparison child welfare branch
offices. Thus, the Safe Start program coordinator was granted access to the
electronic case file database used by child welfare offices to document and record
information about new and existing cases. On a monthly basis, the Safe Start
program coordinator would review all new cases (over the past 30 days) in the
database and record information about potentially eligible cases. This case
information was then passed to designated child welfare personnel to seek
consent to contact from the caregiver. For interested caregivers, contact
information would be passed to the research team, which would in turn contact
the family to explain the study.
Measures
The measures used in this study are described fully in Chapter Two of the
main document (see http://www.rand.org/pubs/technical_reports/
7
TR991-1.html). The measures were uniform across the national evaluation but
prioritized within each site as to the relevance to the intervention under study.
Table 1 displays the prioritization of primary and secondary outcome measures
for Multnomah County’s intervention. As discussed in the following paragraphs,
the enrollment and retention rates were low for this site. Thus, several outcome
measures selected by the site could not ultimately be included because the
number of children assessed on those measures was inadequate to allow for even
descriptive summaries of the measures (i.e., there were fewer than five children
in the group). This included the family involvement measure of the caregiverchild relationship domain (selected as a secondary outcome) and both domains
selected as tertiary outcomes: (1) the school readiness/performance domain
(measured by the Woodcock-Johnson III) and (2) the affective strengths and
school functioning measures of the social-emotional competence domain.
8
Table 1
Prioritized Outcome Measures for Multnomah County Safe Start
Primary Outcome Measures
Domain
Source/Measure
Age of
Child
All
Respondent
Caregiver-Child
Relationship
Violence Exposure
Parenting Stress Index
Juvenile Victimization
Questionnaire
All
Caregiver
Caregiver Victimization
Caregiver Victimization
Questionnaire
All
Caregiver
Secondary Outcome Measures
Domain
Measure
Caregiver
Age of
Child
3–6 years
Respondent
PTSD Symptoms
Trauma Symptom Checklist for
Young Children
Behavior/Conduct
Problems
BITSEA and Behavior Problem
Index
1–6 years
Caregiver
Social-Emotional
Competence
ASQ
0–2 years
Caregiver
Social-Emotional
Competence
BITSEA and SSRS (Assertion
and Self-Control)
1–6 years
Caregiver
Social-Emotional
Competence
Background and
Contextual Factors
SSRS (Cooperation)
3–6 years
Caregiver
All
Caregiver
Everyday Stressors Index
Caregiver
NOTE: ASQ = Ages and Stages Questionnaire, BITSEA = Brief Infant-Toddler Social and
Emotional Assessment, SSRS = Social Skills Rating System.
Enrollment and Retention
The Multnomah County Safe program sought to recruit eligible families
from within a child welfare population. According to data submitted on its
Quarterly Activity Reports, Multnomah County Safe Start enrolled 12 percent of
the families referred to the program. The most common reasons that families
were not enrolled related to legal guardian–related issues, such as inability to
locate or obtain permission from the noncustodial parent (61 percent), and
caregiver-related issues, such as inability to locate or missed appointments (38
percent).
In Table 2, we present the number and percent of all enrollees who were
eligible for participation at each data collection time point. At a minimum, a
9
completed caregiver packet is necessary for a family to be included in the
outcome analyses. The overall retention rate for the study was 37 percent for
caregivers at six months and 29 percent for children. As shown in Table 2, the
retention rate was somewhat lower for intervention group caregivers at six
months, relative to comparison group caregivers. Specifically, the Multnomah
County program staff initially enrolled 32 families in the intervention group and
was able to complete a six-month research assessment with 34 percent of
caregivers. For the initially enrolled 11 comparison group families, the program
retained 45 percent of caregivers at six months. By the 12-month mark, only
seven intervention group caregivers were participating in the research
assessment and only one comparison group caregiver.
Because of low overall study participation, the Multnomah County site
stopped enrollment of new families about two months before the end of
enrollment for the national evaluation and stopped following enrolled families
about eight months before other sites. Difficulties with enrollment and retention
had to do with the nature to the target population (a child welfare site), coupled
with the study design (recruiting into a no-additional-service comparison group),
and limited resources available to supplement data collection (Schultz et al.,
2010).
It must be noted that the low retention rate increases the potential for
biased results in analyses that could be conducted with this sample size. In other
words, this degree of attrition may be related to treatment factors that lead to
selection bias. For example, if families in more distress are more likely to leave
the study at a higher rate, the results can be misleading.
10
Table 2
Retention of Enrollees Eligible to Participate in Assessments at Each Time
Point
Caregiver Assessment
Six
12
18
24
Months Months Months Months
Intervention
Received
11
Expected*
32
Retention
Rate
34%
Comparison
Received
5
Expected*
11
Retention
Rate
45%
Overall
Retention
Rate
37%
Six
Months
Child Assessment
12
18
24
Months Months Months
7
19
2
11
0
5
6
20
4
16
2
10
0
4
37%
18%
0%
30%
25%
20%
0%
1
6
0
2
0
0
1
4
1
4
0
1
0
0
17%
0%
0%
25%
25%
0%
0%
32%
15%
0%
29%
25%
18%
0%
* The number of expected caregiver assessments for longer-term assessments differs from the
number who entered the study because the field period for collecting data in this study ended in
the fall of 2009, before all families entered the window of time for assessments at 12, 18, or 24
months.
Special Issues
In the Multnomah County Safe Start outcomes evaluation, recruitment
and retention of families in a child welfare context was challenging. Many
caregivers were difficult to locate, either initially or after enrollment had
occurred. Moreover, many located caregivers did not provide their consent to be
contacted by the research team. Anecdotally, it appeared that the initial lack of
interest was often due to fear or reluctance to participate in anything offered in a
child welfare context and to the difficult circumstances faced by many caregivers
at the time of the initial child welfare referral. Follow-up assessments were rarely
completed with enrolled families because most either could not be located or did
not demonstrate continuing interest in participation. These recruitment and
retention difficulties were magnified for the comparison group since specialized
Safe Start services were not offered along with study participation. For a more indepth discussion, see Schultz et al. (2010).
11
Analysis Plan
First, we conducted descriptive analyses to summarize the sample
baseline characteristics: age, gender, race or ethnicity, the family income level,
and the child’s violence exposure at baseline. We also compared the two groups
on primary, secondary, and tertiary outcomes at baseline. Because participants
were not randomized to groups, there was an increased likelihood that the
groups may differ in some ways. To assess this possibility, we tested for
differences in child and caregiver characteristics and outcomes between
intervention and comparison group children using t-tests and chi-square tests.
While 43 children were enrolled in the Safe Start program at baseline (32
in the intervention and 11 in the comparison group), only 16 children total were
retained at the six-month follow-up, and only eight children were retained at 12
months. These sample sizes are too small to allow for any estimation of the effect
of the intervention on the participants’ outcomes over time and for examining
difference within the intervention group by the amount of services received.
Without a comparison group, we were only able to conduct limited analyses.
When the sample size allowed, we examined differences between the
intervention and comparison groups at baseline using t-tests. We also examined
changes within the intervention group between baseline and six months using ttests. Because we could not conduct statistical tests comparing the two groups at
the follow-up assessments, these results should be interpreted with caution.
Because the participants were not randomly assigned to groups in this study, it
was not possible to account for any pre-existing group differences.
Even though we conducted just one type of statistical test in these
analyses (changes within the intervention group only), we conducted this test
multiple times because of the large number of primary and secondary outcome
measures. When conducting large numbers of simultaneous hypothesis tests, it is
important to account for the possibility that some results will achieve statistical
significance simply by chance. The use of a traditional 95-percent confidence
interval, for example, will result in one out of 20 comparisons achieving
statistical significance as a result of random error. We therefore adjusted for false
positives using the False Discovery Rate (FDR) method (Benjamini and Hochberg,
1995). Our assessments of statistical significance were based on applying the FDR
procedure separately to all of the primary, secondary, and tertiary outcome tests
12
in this report using an FDR of 0.05. The FDR significance level differed for
unadjusted difference in difference models because the number of statistical tests
varied by outcome type. With the 12 statistical tests conducted for the
intervention group on the primary outcomes, this led to adopting a statistical
significance cutoff of 0.004. With five secondary outcomes tested, the FDR
significance level adopted was 0.01.
In the discussion of results, we have also identified nonsignificant trends
in the data, defined as those tests resulting in p-values of less than 0.05 but not
exceeding the FDR criterion for statistical significance. These trends may suggest
a practical difference that would be statistically significant with a larger sample
size. By the same token, however, they must be interpreted with caution because
we cannot rule out that the difference was due to chance because of the multiple
significance tests being conducted.
RESULTS
Baseline Descriptive Statistics
For the descriptive statistics, we provide the characteristics for the full
enrolled sample at baseline. As shown in Table 3, the baseline sample was
composed of 65 percent female children, with an average age of 3.6 years. The
Multnomah County Safe Start site enrolled children between birth and 6 years of
age. Children ages 3 to 5 made up 49 percent of all enrolled children, and those 2
and younger represented 42 percent of the overall sample. The children in the
sample were predominately Hispanic (40 percent), with white (30 percent)
children making up the second largest group. Black children represented 7
percent of the overall sample, and remaining were other race/ethnicity children
(23 percent). Many respondents did not provide responses to questions about
household income. Of the 23 that did answer, almost 70 percent reported a
family income of less than $30,000 per year, with the intervention group having
lower family incomes than the comparison group (see Table 3). According to the
caregiver reports, children in the baseline sample had been exposed to an
average of 2.9 types of violence in their lives prior to the baseline assessment. All
but one of the caregivers who completed the assessments were the target child’s
mother. As noted in the table, there were no statistical differences for these
13
characteristics between the intervention and comparison groups (where
statistical tests could be conducted).
In the sample of families retained at six months, the demographics were
roughly similar to those at baseline, but with only five families retained in the
comparison group, statistical comparisons could not be conducted (data not
shown).
We also examined the Multnomah County Safe Start sample overall at
baseline on two outcomes (PTSD symptoms and caregiver-child relationship) to
assess the level of severity on these indexes among families entering the study.
As shown in Table 4, at baseline, 23 percent of the full sample reported child
PTSD symptoms that fell in the significant range, whereas 73 percent fell in the
normal range. In terms of the caregiver-child relationship, 33 percent of the full
sample had total stress levels that fell in the clinical range. For the different
subscales, 33 percent of the sample had clinical levels of parental distress, 19
percent had clinical levels of parent-child dysfunctional interaction, and 26
percent had clinical levels of difficult child responses.
14
Table 3
Multnomah Safe Start Sample Characteristics for Families in the Baseline
Assessment Sample
Child Characteristics
Age
CR Violence Exposure
Gender
Male
Female
Race/Ethnicity
Hispanic
White
Black
Other
Caregiver Characteristics
Family Income Level
Less than $5,000
$5,000–$10,000
$10,001–$15,000
$15,001–$20,000
$20,001–$30,000
More than $30,000
Relationship to Child
Parent-Guardian
Other Relationship
Combined
Intervention
Comparison
N
43
43
N
Mean
3.6
2.9
%
N
32
32
N
Mean
3.8
2.8
%
N
11
11
N
Mean
2.9
3.1
%
15
28
34.9
65.1
13
19
40.6
59.4
2
9
18.2
81.8
17
13
3
10
N
39.5
30.2
7.0
23.3
%
15
10
0
7
N
46.9
31.3
0.0
21.9
%
2
3
3
3
N
18.2
27.3
27.3
27.3
%
5
1
4
4
2
7
21.7
4.3
17.4
17.4
8.7
30.4
3
1
3
4
1
3
20.0
6.7
20.0
26.7
6.7
20.0
2
0
1
0
1
4
25.0
0.0
12.5
0.0
12.5
50.0
42
1
97.7
2.3
31
1
96.9
3.1
11
0
100.0
0.0
Test for
Comparison
P-Value
0.18
0.77
0.18
NOTES: CR = Caregiver Report. Percentages may not total 100 percent because of rounding.
15
Table 4
Baseline Assessment Estimates for Multnomah Safe Start Families
CR PTSD Symptoms for Ages 3–10
Normal
Borderline
Significant
CR Total Parenting Stress for All
Ages
Parental Distress—Clinical
Parent-Child Dysfunctional
Interaction—Clinical
Difficult Child—Clinical
Total Stress—Clinical
Combined
N
%
N
Boys
%
N
Girls
%
16
1
5
N
73
5
23
%
5
0
1
N
83
0
17
%
11
1
4
N
69
6
25
%
14
8
33
19
3
2
20
13
11
6
39
21
11
14
26
33
3
4
21
29
8
10
29
36
NOTE: CR = Caregiver Report.
Finally, we examined differences between the intervention and
comparison group at baseline for the Multnomah County program’s primary and
secondary outcomes (see this report’s appendix). Primary outcomes for
Multnomah County were child violence exposure, caregiver victimization, and
parent-child relationship. No statistically significant differences were observed
between the groups overall on these measures (see Table A.1, first column).
Secondary outcomes included caregiver and child PTSD symptoms, child
behavior problems, some aspects of social-emotional competence, caregiver
victimization, caregiver resource problems, and caregiver personal problems.
Again, no statistical differences were observed at baseline between the two
groups on these secondary outcome measures (see Table A.2, first column).
Uptake, Dosage, and Process of Care
Family-level service data were recorded by the program on the follow-up
Family Status Sheet and submitted at six-month intervals following initial
enrollment (see Chapter Two of the main document [http://www.rand.org/
pubs/technical_reports/TR991-1.html] for a description). Tables 5a and 5b show
the type and amount of services received by the families enrolled in the
intervention group. The program contained three central components: advocacy,
case coordination, and CPP therapy, but parent support groups were added to
the program as a secondary service.
Table 5a presents the results for services received for all families who were
initially enrolled in the intervention group, regardless of whether they continued
16
to participate in the ongoing research assessment. The data displayed in Table 5a
include services received by summing all time points reported by the program,
with a maximum of 24 months of service provision. Service data for at least one
follow-up wave were available on all 32 initially enrolled intervention group
families. As shown in Table 5a, 63 percent of the intervention group families
received at least one session with a Multnomah County advocate. While the
average was 20 sessions, 45 percent of the full intervention group received ten or
fewer sessions. The program reported the reason that advocacy services ended
for 14 of the 20 families that received this service. The majority (64 percent)
ended because of successful completion, with the remaining ending because the
family dropped out or could not otherwise continue (three, 21 percent), or for
some other reason (two, 14 percent). Case coordination meetings also took place
for 63 percent of the full baseline sample, with an average of eight meetings for
families receiving this service. The most common reason that case coordination
meetings ended was successful completion (43 percent of 16 families with
reported data). Families otherwise discontinued services in 31 percent of cases,
with the remaining 25 percent (four of 16) of case coordination meetings ending
for other reasons.
CPP therapy was provided to 28 percent of the intervention group
families, with an average number of 16 sessions for the group. The program
reported the reason for therapy sessions ending for five of the nine families
receiving this service. In three cases the family dropped out of therapy, with the
remaining two families successfully completing the program. Parent support
groups were attended by five (16 percent) of the intervention group, with 60
percent of those attending between one and ten group sessions. Of these five
participating families, the program reported that three successfully completed
the parent support groups, while two families dropped out prior to completion.
17
Table 5a
Services Received by Multnomah County Safe Start Intervention Families
(Baseline Sample)
Service
Number
with
Service
20
Percentage
with
Service*
63%
CPP Therapy
9
28%
1–33
Case
Coordination
20
63%
1–53
Parent Support
Groups
5
16%
2–20
Advocacy
Range
1–129
Distribution
1–5
6–10
11–20
21+
1–5
6–10
11–20
21+
1–5
6–10
11–20
21+
1–5
6–10
11–20
21+
25%
20%
30%
25%
33%
11%
11%
44%
80%
5%
10%
5%
20%
40%
40%
0%
Mean
Median
20
12
16
13
8
1
10
8
*
The denominator is the 32 intervention group families who were initially enrolled in the
intervention group for whom one or more follow-up Family Status Sheets were submitted.
NOTE: Percentages may not total 100 percent because of rounding.
Table 5b shows the services received by the subgroup of intervention
group families who participated in the six-month follow-up research assessment.
These are the 11 families included in the intervention group in the outcome
analyses sample for the Multnomah County program. Table 5b shows the
services they received within the six-month period between baseline and the sixmonth assessment. Eight (73 percent) of the 11 families in the outcome analyses
sample received at least one session with a Safe Start advocate, with an average
of 23 sessions. The number of sessions received ranged from three to 99, with the
majority receiving between 11 and 20 sessions. The program reported that fivw
of the eight participating families successfully completed their advocacy services
but reported no data on the remainder. Of the eight families who received case
coordination services, the majority of cases (80 percent) involved between one
and five meetings. The program reported case coordination ending successful for
three of the eight families, but they were ended for other reasons in two cases,
and no data were reported for the remaining three families.
As shown in Table 5b, just three (27 percent) of the 11 analysis group
families participated in the CPP therapy service, and the program did not report
18
why therapy services ended, after an average of 13 sessions for families in this
group. Two (18 percent) participated in the parent support group service, with
both reported to have completed this service.
Table 5b
Six-Month Services Received by Multnomah County Safe Start Intervention
Families (Six-Month Analysis Sample)
Service
Number
with
Service
8
Percentage
with
Service*
73%
Range
CPP Therapy
3
27%
2–19
Case
Coordination
8
73%
1–37
Parent Support
Groups
2
18%
9–20
Advocacy
3–99
Distribution
1–5
6–10
11–20
21+
1–5
6–10
11–20
21+
1–5
6–10
11–20
21+
1–5
6–10
11–20
21+
13%
13%
63%
13%
33%
0%
66%
0%
80%
5%
10%
5%
0%
50%
50%
0%
Mean
Median
23
13
13
10
8
2
15
9
*
The denominator is the 11 intervention group families with a follow-up Family Status Sheet
at the six-month assessment point who participated in the six-month research assessment.
NOTE: Percentages may not total 100 percent because of rounding.
Outcomes Analysis
As previously discussed, inadequate data substantially limited the
analysis of data for this site. As such, we were not able to conduct any statistical
comparisons between groups. Mean differences over time could be tested only
for the intervention group and only through the six-month assessment point.
Mean Differences over Time
Table 6 shows differences over time for the Multnomah County program’s
primary outcomes, comparing changes for each individual intervention group
family between baseline and six months when the sample size allowed. For
primary outcomes, there was one statistically significant difference within the
intervention group, with a significant decline in reports of domestic violence for
between baseline and six months. This finding is expected, given the different
19
reference period for the baseline and follow-up assessments. Specifically, the
baseline assessment asks caregivers about victimization over the prior 12 months,
while the follow-up assessments ask about the prior six months. Table 6 also
shows a nonsignificant trend toward declines in the child witnessing violence.
Likewise, this finding is expected, as these measures compare lifetime exposure
at baseline to exposure over a six-month period at the follow-up assessment.
Table 7 shows changes over time on Multnomah County’s secondary
outcomes for the intervention group, except for measures of child PTSD
symptoms and child cooperation, as there were not adequate data to test withinfamily changes. Changes were generally in the direction expected, but for those
with adequate sample size none of the differences reached statistical significance.
20
Table 6
Changes in Intervention Group Means for Primary
Outcome Variables Between Baseline and Six-Month
Assessment
Primary Outcome
Caregiver-Child Relationship
CR Parental Distress for Ages 0–12
CR Parent-Child Dysfunction for Ages 0–
12
CR Difficult Child for Ages 0–12
CR Total Parental Stress for Ages 0–12
Violence Exposure
CR Total Child Victimization Experiences
for Ages 0–12
CR Child Maltreatment for Ages 0–12
CR Child Assault for Ages 0–12
CR Child Sexual Abuse for Ages 0–12
CR Child Witnessing Violence for Ages
0–12
CR Caregiver Total Number of Traumatic
Experiences
CR Caregiver Experience of Any Non-DV
Traumasb
CR Caregiver Experience of Any
Domestic Violenceb
N
WithinFamily Mean
Changesa
11
1.45
1.91
11
11
11
2.82
6.18
11
–1.36
11
11
11
–0.27
0.18
–0.09
–1.20
#
10
–0.18
11
–0.09
11
11
a
–0.55
*
This column reflects within-family mean changes between the baseline and six-month scores for
each group separately. * indicates a significant paired t-test of differences over time.
b
This outcome is a categorical variable, and the unadjusted within-family mean change is a
change in proportion.
NOTES: CR = Caregiver Report; DV = domestic violence. # indicates a nonsignificant trend in the
t-test (p<0.05 but does not meet the FDR correction threshold). Mean change estimates are not
shown when the group size is fewer than ten, and comparisons are not shown when the group
size is fewer than ten for either group.
21
Table 7
Changes in Intervention Group Means for Secondary Outcome Variables
Between Baseline and Six-month Assessment
Secondary Outcome
Background and Contextual Factors
CR Caregiver Resource Problems
CR Caregiver Personal Problems
Behavior/Conduct Problems
CR Child Behavior Problems for Ages 1–
18
Social-Emotional Competence
CR Child Assertion for Ages 1–12
CR Child Self-Control for Ages 1–12
N
WithinFamily
Mean
Changesa
11
11
–0.91
0.73
10
0.04
10
10
–0.28
0.15
a
This column reflects within-family mean changes between the baseline and six-month scores for
each group separately. * indicates a significant paired t-test of differences over time.
NOTES: CR = Caregiver Report. Mean change estimates are not shown when the group size is
fewer than ten, and comparisons are not shown when the group size is fewer than ten for either
group.
CONCLUSIONS
The Multnomah County Safe Start program involved three main
components: domestic violence advocacy, a modified version of CPP, and case
coordination and consultation. The planned evaluation was a quasi-experimental
trial enrolling similar children in different child welfare offices within
Multnomah County. The Safe Start program staff enrolled 43 families in the
study, 32 in the intervention group and 11 in the comparison. By the time of the
first follow-up assessment, only 11 families remained in the Multnomah County
intervention group and five in the comparison group. Consequently, evaluation
of the impact of the intervention was not possible because of an inability to
compare groups on post-intervention outcomes.
Because of low overall study participation, the Multnomah County site
stopped enrollment of new families about two months before the end of
enrollment for the national evaluation and stopped following enrolled families
about eight months before other sites. Difficulties with enrollment and retention
had to do with the nature to the target population (a child welfare site), coupled
with the study design (recruiting into a no-additional-service comparison group)
and limited resources available to supplement data collection.
22
The baseline data suggest that the overall sample had been exposed to an
average of 2.9 types of violence in their lives prior to the baseline assessment.
Caregivers reported significant PTSD symptoms for 23 percent of children in the
overall sample, and 33 percent reported total parenting stress levels in the
clinical range. Multnomah County’s tailored approach to services meant that
families in the intervention group received different types of services, depending
on the circumstances. Overall, 63 percent of the intervention group families
received at least one session with a Multnomah County advocate and received at
least one case coordination meeting. CPP therapy was provided to 28 percent of
the intervention group families, with an average number of 16 sessions for the
group. Of the 11 intervention group families in the retained sample at six months,
eight (73 percent) received at least one session with a Safe Start advocate, with an
average of 23 sessions. Eight families who received case coordination services
and three families in the retained sample participated in at least one CPP therapy
service.
Limited statistical testing was conducted for six-month changes within the
intervention group only, but no significant changes in outcomes were observed,
with the exception of caregiver experience of domestic violence. This change
would be expected, however, because of different reference periods for the
baseline and follow-up assessments.
Overall, the sample size limitations mean that no conclusions can be
drawn about the impact of the Multnomah County Safe Start intervention childand family-level outcomes. However, our process evaluation found that the
program model did appear to have important impacts in terms of increasing
integration and coordination of services between child welfare and domestic
violence advocacy service providers (Schultz et al., 2010). This is a positive
finding, as system-level change was a key goal of the program. The intervention
model’s impact on individual outcomes, however, remains unknown, and
conclusions await testing with an adequate research sample.
23
REFERENCES
Aron, L. Y., and K. K. Olson, “Efforts by Child Welfare Agencies to Address Domestic
Violence: The Experiences of Five Communities, Washington, D.C.: The Urban
Institute, 1997.
Benjamini, Y., and Y. Hochberg, “Controlling the False Discovery Rate: A
Practical and Powerful Approach to Multiple Testing,” Journal of the Royal
Statistical Society, Series B, Vol. 57, 1995, pp. 289–300.
Lieberman, A.F., and P. Van Horn, “Don’t Hit My Mommy!” A Manual for ChildParent Psychotherapy with Young Witnesses of Family Violence, Washington,
D.C.: Zero to Three, 2005.
Lieberman, A. F., P. Van Horn, and C. Ghosh Ippen, “Toward Evidence-Based
Treatment: Child-Parent Psychotherapy with Preschoolers Exposed to
Marital Violence,” Journal of the American Academy of Child and Adolescent
Psychiatry, Vol. 44, 2005, pp. 72–79.
Multnomah County, Funding Proposal to the Office of Juvenile Justice and
Delinquency Prevention—CFDA Title: Safe Start: Promising Approaches for
Children Exposed to Violence, Multnomah County, Oreg., 2004.
Schechter, S., Model Initiatives Linking Domestic Violence and Child Welfare, Iowa
City, Ia.: University of Iowa, School of Social Work, prepared for the
conference on Domestic Violence and Child Welfare: Integrating Policy
and Practice for Families, Racine, Wis., June 8–10, 1994.
Schultz, D., L. H. Jaycox, L. J. Hickman, A. Chandra, D. Barnes-Proby, J. Acosta,
A. Beckman, T. Francois, and L. Honess-Morealle, National Evaluation of
Safe Start Promising Approaches: Assessing Program Implementation, Santa
Monica, Calif.: RAND Corporation, TR-750-DOJ, 2010. As of July 17, 2011:
http://www.rand.org/pubs/technical_reports/TR750.html
24
MULTNOMAH OUTCOMES APPENDIX
25
Table A.1
Comparison of Means for Primary Outcome Variables over Time
Primary Outcome
Caregiver-Child Relationship
CR Parent Distress for
Intervention
Ages 0–12
Comparison
CR Parent-Child
Intervention
Dysfunction for Ages 0–12 Comparison
CR Difficult Child for Ages Intervention
0–12
Comparison
CR Total Parenting Stress
Intervention
for Ages 0–12
Comparison
Violence Exposure
CR Total Child
Intervention
Victimization Experiences
Comparison
for Ages 0–12
CR Child Maltreatment for Intervention
Ages 0–12
Comparison
CR Child Assault for Ages Intervention
0–12
Comparison
CR Child Sexual Abuse for Intervention
Ages 0–12
Comparison
CR Child Witnessing
Intervention
Violence for Ages 0–12
Comparison
CR Caregiver Total
Intervention
Number of Traumatic
Comparison
Experiences
CR Caregiver Experience
Intervention
of Any Non-DV Trauma
Comparison
N
Baseline
Mean
Six Months
N
Mean
N
32
11
32
11
32
11
32
11
31.00
29.64
22.56
19.45
29.13
25.64
82.69
74.73
11
5
11
5
11
5
11
5
30.27
27.80
21.18
16.00
28.45
22.60
79.91
66.40
7
1
7
1
7
1
7
1
25.29
32
11
2.84
3.09
11
5
1.73
0.20
7
1
0.71
32
11
32
11
32
10
31
8
32
11
0.72
0.64
0.34
0.45
0.06
0.00
1.65
2.00
0.41
0.18
11
5
11
5
11
5
11
4
11
5
0.64
0.00
0.55
0.00
0.00
0.00
0.55
0.14
0.09
0.20
7
1
7
1
7
1
7
1
7
1
32
11
0.16
0.27
11
5
0.09
0.00
7
1
0.00
26
#
12 Months
Mean
19.00
27.14
71.43
0.29
0.00
0.14
0.00
Table A.1—continued
Primary Outcome
CR Caregiver Experience
of Any DV
Intervention
Comparison
N
32
11
Baseline
Mean
0.75
0.73
Six Months
N
Mean
11
0.18
5
0.00
N
7
1
12 Months
Mean
0.14
NOTES: CR = Caregiver Report; DV = domestic violence. # indicates nonsignificant trend (p<0.05 and > FDR
significance criterion). * indicates statistically significant (p-value<FDR significance criterion). Data are not shown for
outcomes when the cell size is fewer than five for the group. Comparisons were not tested when the group size was
fewer than ten for either group.
27
Table A.2
Comparison of Means for Secondary Outcome Variables over Time Secondary Outcome
Background and Contextual Factors
CR Caregiver Resource
Intervention
Problems
Comparison
CR Caregiver Personal
Intervention
Problems
Comparison
PTSD Symptoms
CR Child PTSD Symptoms Intervention
for Ages 3–10
Comparison
Behavior/Conduct Problems
CR Child Behavior
Intervention
Problems for Ages 1–18
Comparison
Social-Emotional Competence
CR Child Assertion for
Intervention
Ages 1–12
Comparison
CR Child Self-Control for
Intervention
Ages 1–12
Comparison
CR Child Cooperation for
Intervention
Ages 3–12
Comparison
CR Family Involvement
Intervention
for Ages 6–12
Comparison
N
Baseline
Mean
Six Months
N
Mean
N
32
11
32
11
15.66
16.55
23.44
25.45
11
5
11
5
14.36
14.60
22.18
21.40
7
1
7
1
20
4
38.60
6
1
35.67
3
1
31
9
–0.14
–0.04
10
4
–0.21
7
1
–0.42
31
9
31
9
18
4
5
2
0.07
0.10
0.26
0.11
12.28
10
4
10
4
5
1
2
0
0.01
7
1
7
1
4
1
2
0
0.08
23.20
0.49
11.80
12 Months
Mean
15.00
22.71
0.19
NOTES: CR = Caregiver Report. # indicates nonsignificant trend (p<0.05 and > FDR significance criterion). * indicates statistically significant (pvalue<FDR significance criterion). Data are not shown for outcomes when the cell size is fewer than five for the group. Comparisons were not
tested when the group size was fewer than ten for either group.
28
Download